Jackson Mark. Machining with Abrasives
Подождите немного. Документ загружается.
8.3.2.2 Polishing Media
Three types of abrasives, silicon carbide (SiC), alumina (Al
2
O
3
) and garnet with
mesh sizes of 100, 200 and 400 were used to examine the effect of abrasive type
and mesh size on the material removal and surface finish. Focusing on the use of
conventional abrasives is based on the consideration of the production cost. The
effects of fine diamond powders are also examined, particularly for fine polishing.
The diamond powder has particle sizes of 40–80 mm and particles in diamond
suspension have a size of 90 mm.
8.3.2.3 Effects of Polishing Conditions
As shown in Table 8.7, the normalized removal weight for the planetary mill using
SiC abrasive of mesh size of 100 is several times higher than that obtained
using the tumbling mill, and more than 20 times greater that with the vibratory
mill. Here the normalized removal weight is defined as the removal amount divided
by the component weight before polishing. The planetary mill demonstrated to be a
more efficient polishing than the tumbling mill due to its higher energy level of
collisions. The tumbling and vibratory mills though have much smaller removal
rates, but both have potential for mass production because the collisions between
components during polishing can be avoided due to their movement styles.
The removal effect involved in the planetary mill is not uniformly distributed.
The removal rate at the egg ends is smaller than that in the middle. However, though
copper egg
abrasives
A
O
w
p
w
v
R
p
R
v
Fig. 8.33 Schematic
illustration of a planetary mill
abrasives
copper egg
Fig. 8.34 Illustration of
motion in a vibratory mill
8 Polishing Using Flexible Abrasive Tools and Loose Abrasives 375
the tumbling mill is not as efficient as the planetary, it is easy to control the removal
rate and has a uniform removal around the free-form component. A higher effi-
ciency with tumbling mill could be achieved by increasing the rotation speed and
size of the tumbler as the current speed is only 136 m/min (for a container diameter
of 90 mm).
R
a
and R
max
are used to characterize the surface finish of the polished egg
samples, measured using a Talor-Hobson (Talysurf) Profilometer. The roughness
values of R
a
and R
max
are 1.87 and 15.85 mm, respectively for the as-received egg
components after belt polishing. In Table 8.7, the values of R
a
and R
max
are much
improved after tumbling 1 or 2 days. The effect of mill type on the surface
roughness is not significant.
The abrasives used also plays an important role in the polishing of free-form
components. As can be seen from Table 8.7, in the tumbling mill for the same mesh
size, alumina has obvious advantage over silicon carbide in terms of removal rate,
though silicon carbide does produce a slightly better surface finish. The combina-
tion of alumina and diamond appears to give a much improved material removal
rate, in comparison to alumina alone. However, during dry polishing in the plane-
tary mill, the surface finish is not much improved. This could be due to aggressive
removal caused by diamond abrasives. The addition of isopropyl alcohol (80 vol.%)
into the combination polishing media (20 wt.%) has significantly improved the
surface finish (whose value of 1.01 is almost the same as that produced by
SiC#100), while still maintaining a reasonabl y high removal rate.
Free abrasives of silicon carbide and alumina as the polishing media in the
tumbling mill are compared. Figure 8.35 shows the normalized removal weights of
SiC and Al
2
O
3
abrasives. It is apparent that the removal rate with Al
2
O
3
abrasives is
much higher than that with SiC.
In Fig. 8.36, values of average surface roughness polished using SiC and alumina
abrasives are plotted as a function of the polishing time. For both the abrasives used
surface roughness is improved with the progress of polishing. When polished for
Table 8.7 Effect of milling conditions on removal weight and surface roughness
Mill type Abrasive type
Polishing
time (h)
Normalized
removal
weight
R
a
(mm)
R
max
(mm)
Tumbling Alumina #100 48 0.24 0.97 9.07
Tumbling SiC #100 48 0.08 0.88 6.44
Planetary SiC #100 48 0.55 1.02 7.23
Planetary Alumina #400
+ Diamond
(40–80 mm)
24 0.98 1.59 11.02
Planetary Alumina #400
+ Diamond
(40–80 mm), wet
24 0.82 1.01 10.97
Vibrotary SiC #100 48 0.02 0.95 8.74
376 H. Huang et al.
27 h, the roughness produced by alumina is lower than that by SiC. However, further
polishing of another 21 h demonstrated that SiC can produce a slightly better surface
finish than alumina. This is because that after 27 h of polishing, uneven spots on the
workpiece surface was still not fully removed. As alumina abrasive produced a
greater removal, the surface finish of the component was thus slightly better. Once
the uneven spots were removed fully, SiC abrasives gives a better surface finish as
they produced a less aggressive removal than alumina.
For all the abrasives the profile of the polished free-form components is much
improved. The polished components clearly have a much smoother surface than
that prior to polishing, as illustrated in Fig. 8.37. The type of polishing machines
determines the polishing efficiency, which is dependent on the relative movements
between the components and the polishing media. The tumbling method is feasible
to replace the complete manual polishing and partial buffing. Nevertheless, a
comprehensive study has to be carried out to find out the optimum combination
of abrasives. In addition, under current tumbling conditions using the tested
abrasives, mirror surface finish won’t be achieved. The development of suitable
polishing media are needed.
0
0.1
0.2
0.3
0.4
0.5
0.6
20 30 40 50 60 70 80
SiC, #100
Al
2
O
3
, #100
Normalised Removal Weight
Tumblin
g
Time (hours)
Fig. 8.35 Removal weight
plotted as a function of
polishing time during
tumbling (container speed
136 m/min, container
diameter 90 mm and filling
ratio 0.8 kg/l volume)
0
1
2
3
020406080
SiC #100
Al
2
O
3
#100
Roughness, R
a
(
µ
m)
Tumbling Time (hours)
Fig. 8.36 Surface roughness
of egg components plotted as
a function of tumbling time
for Al
2
O
3
#100 and SiC #100
abrasives (container speed
136 m/min, container
diameter 90 mm and filling
ratio 0.8 kg/l volume)
8 Polishing Using Flexible Abrasive Tools and Loose Abrasives 377
8.3.2.4 Effect of Tumbling on Buffing Cycle Time
To evaluate the effect of tumbling on the buffing cycle time, five egg components
polished using the tumbling method with alumina #100 abrasives were further
buffed manually. The buffing times were then recorded and are compared with
the average cycle times from the production line (shaded numbers) in Table 8.8.
It is seen from Table 8.8 that second stage of belt polishing can be spared and the
buffing time can also be shortened, saving about 30% of the total polishing time.
Note that here the tumbling time is not calculated as tumbling can be operated in
mass production, not requiring skilled workers. Figure 8.38 shows the surfaces after
manual belt polishing and after polishing using tumbling method. The tumbling
using Al
2
O
3
abrasives improves both the surface fin ish and egg profile, thus, it is not
surprising that the buffing time had been shortened. It should be noted that the fine
buffing time for Part 5 was significantly longer than the industrial cycle time
because the rough buffing was removed from the processing.
Table 8.8 Comparison of
cycle times (in minutes) with
and without tumbling
a
Part
number
Belt
polishing
(#150)
Belt
polishing
(#320)
Rough
buffing
+
Fine
buffing
+
Time
improved
(%)
Manual 3 7 3 12.5
13 3 10 37.3
23 3 12 29.5
33 3 12 29.5
43 3 13 25.5
53 15 29.5
+ sign indicates that the process was NOT applied
a
Buffing was carried out by a skilled worker in the production line
Fig. 8.37 Surfaces of egg component (a) prior to machining and (b) after polishing using the
tumbling method with Al
2
O
3
#100
378 H. Huang et al.
The results of tumbling experiments can be summarized as follow s
l
The large lumps on the egg surface could not be removed using the tumbling
method due to its uniform polishing effect. So a rough polishing with abrasive
belt was still required as the lumps had to be removed manually.
l
The egg components polished using the tumbling method had much smoother
profiles than those before tumbling. However, the surface finish was not signifi-
cantly improved. It is apparent that the surface finish strongly depends on the
polishing media used. Optimum combinations of polishing media, including
the combination of different abrasives or mesh sizes, should be further studied
to achieve a satisfactory surface finish.
l
Using Al
2
O
3
abrasives produced a more efficient polishing than SiC abrasives,
but using SiC abrasives resulted in a slightly better surface finish.
l
The planetary mill is the most efficient polishing machine, followed by the
tumbling mill, and then vibratory mill. But the tumbling mill enables the most
uniform material removal across the exposed surface.
l
The tumbling method is feasible to replace fine belt polishing and rough buffing.
Without inclusion of the tumbling time, the total machining cycle time was
about 30% reduced.
8.3.2.5 Barrel Polishing
The results from the tumbling experiments indicated that polishing media is an
important factor which controls the surface finish of polished products. Polishing
was further conducted using different polishing media, which are commercially
available from a barrel polishing specialist in Japan, UJIDEN. Two types of
polishing media were tested, which were specially developed for polishing copper
egg components. The barrel polishing experiment was carried out using a UJIDEN
centrifugal barrel, whose working principle is similar to the tumbling or planetary
mill, but with high energy level. The barrel finishing had two steps: rough and fine
polishing using two different types of the barrels, respectively. As can be seen in
Fig. 8.38 Egg component (a) after manual belt polishing, (b) after tumbling using Al
2
O
3
abra-
sives, and (c) after manual buffing
8 Polishing Using Flexible Abrasive Tools and Loose Abrasives 379
Table 8.9, TOSALIT PC-6-15 was used for rough polishing, which combined
conventional abrasives with a cleaning/cooling compound (NF-44). The NF-44
compound was used to reduce the scratching from the work material being removed
and prevent from the over-heating of egg component s during polishing. For fine
polishing two different combinations, i.e. WL#46 + AM-50 and WL#36 + AM-55,
were used to examine the effect of mesh size of walnut particles on the surface
roughness. Note WL#36 and WL#46 are polishing materials made of walnut shells,
a non-conventional polishing media. Again, AM-55 is a cooling media. In the fine
polishing, rubber containers were used to avoid collisions, which had the similar
function to the plastic container used in our tumbling experiments. Detailed polish-
ing conditions are listed in Table 8.9 (Fig. 8.39 ).
The polishing results were satisfactory after a total of 3 h (rough + fine) polishing.
Figure 8.38 shows two copper eggs polished using the barrel finishing method with a
mirror surface finish. It was difficult to visually distinguish the difference of surface
roughness produced by the barrel polishing and manual buffing (Fig. 8.37c).
8.4 Concluding Remarks
The polishing processes with flexible abrasive tools, including belts and films, free
abrasives and abrasive slurries have been investigated. The polished products are of
very different geometries, varied from three-dimensional free-form surfaces to
inner wall surfaces of micro bores. Very different work materials were also studied,
including difficult-to-machine superalloys, hard and brittle solids and soft and
ductile metals.
The development of the robotic polishing processes for automation of the
manual operations of aerospace engine components, which are labour intensive
and requires extensive process knowledge and polishing skills, involved a great
deal of investigat ions of polishing hardware and optimal process parameters.
Aerospace engine components are made of superalloys, which fall into difficult-
to-machine category. Benchmark tests showed that the systems achieved better
Table 8.9 Conditions and results from barrel polishing of egg components
Process Rough polishing Fine Polishing
Type Planetary Planetary
Barrel machine Capacity
4 rooms (6 L/room)
2–3 eggs/room
160 rubber rooms (2.6 L/room)
1 egg/room
Rotation speed 140 rpm 120 rpm
Polishing media TOSALIT PC-6-15 TOSALIT WL #46 TOSALIT WL #36
Filling ratio (work/media) (%) 50 50 50
Cleaning/cooling compound NF-44 AM-50 AM-55
Polishing time (h) 2 1 1
Roughness, R
a
(nm) ~22 ~30
Egg profile Perfect Perfect
380 H. Huang et al.
and more consistent quality and considerably reduced the cycle time s, in comparison
to the respective manual operations. In particular, for the polishing system for
free-form turbine airfoils, it demands for the synergistic combination of process
and system solutions to meet the stringent quality requirement and 3D design
criteria, such as profile smoothness and airfoil wall thickness. Though the surface
roughness required for the components is not so high as the conventional polishing
deals with, the technologies developed has great implications to the automation of
finishing processes for precision mechanical components, such as three-dimensional
moulds or dies.
The polishing of optic fibre connectors produced mi rror surface finish and
stringent geomet ric accuracies. Since the connec tors consist of two different mate-
rials, i.e. zirconia and silica, the development of the polishing process for them thus
involved a great deal of investigations of optimal selection of polishing abrasives
and suspensions. There is a distinct difference in material response to polishing
between glass fibre and zirconia ferrule, especially with coarser abrasives in which
brittle chipping occurs in silica fiber, but plastic deformation dominates in the
removal of zirconia material. For coarse polishing, diamond abrasives are not
recommended because they induce severe surface and subsurface damage in silica.
Silicon carbide abrasives can be applied to have efficient material removal to
reduce the fiber height with only moderate surf ace and subsurface damage in
fiber. For fine polishing, fine diamond abrasives produced damage-free and smooth
end faces, scaling average surface roughness smaller than 20 and 30 nm for fiber
and ferrule, respectively. The results also showed that the suspensions have a
significant effect on the geometrical quality and the optical performance of the
polished connectors. The colloidal silica suspe nsion resulted in a markedly
improved geometrical quality. The optimal combination of polishing media and
suspension enables the improvement in cycle time, while maint aining the satisfac-
tory quality requirements.
In the abrasive flow polishing of micro bores, the results showed that the polishing
surface was significantly influenced by the bore size being polished when both the
Fig. 8.39 Copper egg components after barrel polishing
8 Polishing Using Flexible Abrasive Tools and Loose Abrasives 381
abrasive flow pressure and the flow volume passing the bore were maintained
unchanged. The average surface finishes of the as received microbores made on
stainless steel, mild steel and zirconia, varied initially, but after several passes of
polishing they reached almost the same value. This indicates that the polishing with
abrasive flow has an efficient removal of uneven parts, such as spots and scratches, of
the work material, which were left from the previous machining procedure, but is
difficult to remove the bulk material. For the polishing conditions used in this
investigation, the final roughness of wall surfaces for those materials reached
0.6 mm. This suggests that the abrasive flow polishing has great potential for
finishing of both metal and ceramic microbores.
The surface quality and efficiency of polishing free-form copper components
with free abrasives are strongly influenced by the abrasives used and the energy
level of the tumbling barrels, less relevant to the barrel types. A couple of important
factors require great attention. First, during polishing, abrasives in the polishing
media are needed to remove the uneven areas of components, and additives are also
required to wrap the removed chips to avoid their damage of polished surface.
In the fine polishing procedure, additives also plays a role in cooling the polished
areas. Second, great care must be taken to avoid the collisions betwee n the compo-
nents themselves and the component and the barrel container. While traditional
abrasives, such as alumina and diamond, still play important role in free abrasive
polishing, non-conventional materials, such as walnut shell, can also be used for
fine polishing.
Acknowledgements The authors are grateful to experimental assistances and valuable discussion
from ZM Gong, XQ Chen, K Ramesh, QF Li, ZJ Xiong, FZ Fang, YC Liu and PL Teo. This work
was financially supported by the Australian Research Council (ARC), the Department of Educa-
tion, Science and Technology (DEST) Australia and the Agency for Science, Technology and
Research (A-STAR), Singapore.
References
1. Kasai T et al (1990) Improvement of conventional polishing conditions for obtaining super
smooth surfaces of glass and metal works. Ann CIRP 39/1:321–324
2. Weule H and Timmermann S (1990) Automation of the surface finishing in the manufacturing
of dies and molds. Ann CIRP 39/1:299–303
3. Evans CJ et al (1998) Rapidly renewable lap: theory and practice. Ann CIRP 47:239–244
4. Cai GO et al (1998) Analysis on lapping and polishing pressure distribution. Ann CIRP
47/1:235–238
5. Spur G and Engel H (1999) Lapping and polishing of glass and ceramics. In: Jahnamir S et al
(ed) Machining of Ceramics and Composites. Marcel Dekker, New York. 177–207
6. Woodbury RS (1958) History of the grinding machine. Technology Press, Boston
7. Stephen TM et al (1987) Control of tool/workpiece contact force with application to robotic
deburring. IEEE J Robot Autom RA-3:7–18
8. Proctor FM and Murphy KN (1989) Advanced deburring system technology. ASME PED
38:1–12
382 H. Huang et al.
9. Bone GM et al (1991) Force control for robotic deburring. ASME J Dyn Syst Meas Control
113:395–400
10. Elbestawi MA et al (1992) An automated planning, control and inspection system for robotic
deburring. Ann CIRP 41/1:397–401
11. Hollowell R and Guile R (1987) An analysis of robotic chamfering and deburring, ASME
DSC 6:73–79
12. Ge DF at al (1995) Dexterous polishing of overhanging sculptured surfaces with a 6-axis
control robot. Proceedings of the IEEE International Conference on Robotics Automation.
Austin, Texas, USA. 2090–2095
13. Ozaki F et al (1995) A force controlled finishing robot system with a task-directed robot
language. J Robot Mechatron 7:383–388
14. Zhou L and Huang H (2002) Automatic robotic system for jet engine overhaul – system design
and development for honeycomb repair. Int J Adv Manuf Technol 19:370–376
15. Zhou L et al (2000) Automatic robotic system for jet engine overhaul – process development
and enhancement for honeycomb repair. J Jpn Soc Precis Eng 66:1895–1900
16. Huang H et al (2002) Robotic grinding/polishing for turbine vane overhaul. J Mater Process
Technol 127:140–145
17. Huang H et al (2003) SMART robotic system for 3D profile turbine vane repair. J Adv Manuf
Technol 21: 275–283
18. Chen XQ et al (2002) Process development and approach for 3D profile grinding/polishing.
In Chen XQ et al (ed) Advanced Automation Techniques in Adaptive Materials Processing.
World Scientific, Singapore. 19–54
19. Chen XQ et al (2002) Adaptive robotic system for 3D profile grinding/polishing. In Chen XQ
et al (ed) Advanced Automation Techniques in Adaptive Materials Processing. World
Scientific, Singapore. 55–90
20. Taguchi G (1986) Introduction to Quality Engineering. Asian Production Organization,
Tokyo
21. Derickson D (1998) Fiber optic test and measurement. Prentice-Hall, New Jersey. 621–638
22. TIA Standard. Telecommunications Industry Association, Arlington
23. Ultra Tec Manufacturing Inc. (2001) Fiber Optic Connector Polishing. Santa Ana
24. Kanda T et al (1995) New micro-finish surface technology for the fabrication of optical device
endfaces. Proc SPIE 2576:84–91
25. Karaki-Doy T et al (1998) Development of a new automatic processing machine for optic-
fiber connector ends. Bull Jpn Soc Precis Eng 22:216–222
26. Huang H et al (2004) Micro/meso ultra precision grinding of fiber optic connectors and its
influence on optic performance. Precis Eng 28:95–105
27. Huang H et al (2004) Ultraprecision abrasive machining of fiber optic connectors. Key Eng
Mater 257/258:171–176
28. Huang H and Yin L (2006) An experimental investigation of fabrication mechanisms of optic
fibre end faces using nano/micro-indentation and nanogrinding. Int J Nanomanuf 1:47–61
29. Yin L et al (2004) Polishing of fiber connectors. Int J Mach Tools Manuf 44:659–668
30. Lin Y and Huang H (2008) Brittle materials in nano-abrasive fabrication of optical mirror-
surfaces. Precis Eng 32:336–341
31. Chen WK et al (2005) Loose abrasive truing and dressing of resin bond diamond wheels for
grinding fibre optic connectors. J Mater Process Technol 159:229–239
32. Yin L et al (2005) Influences of nanoscale abrasive suspensions on the polishing of fibre optic
connectors. Int J Adv Manuf Technol 25:685–690
33. Yin L et al (2004) High-quality grinding of polycrystalline silicon carbide spherical surfaces.
Wear 256:197–207
34. Masuzawa T (2000) State of art of micromachining. Ann CIRP 49:473–488
35. Van Brussel H et al (2000) Assembly of Microsystems. Ann CIRP 49:451–472
36. Huang H and Kuriyagawa T (2007) Nanometric grinding of axisymmetric aspherical mould
inserts for optic/photonic applications. Int J Machining Machinability Mater 2:71–84
8 Polishing Using Flexible Abrasive Tools and Loose Abrasives 383
37. Chen WK et al (2005) Machining of micro aspherical mould inserts. Precis Eng 29:315–323
38. Ramesh K et al (2004) Microgrinding of deep micro grooves with high table reversal speed.
Int J Mach Tools Manuf 44:39–49
39. Masuzawa T and Tonshoff HK (1997) Three-dimensional micromachining by machine tools.
Ann CIRP 46/2:621–628
40. Huang H et al (2003) Ultrasonic vibration assisted electro-discharge machining of microholes
in Nitinol. J Micromech Microeng 13:693–700
41. Odom B (2001) Making microholes. Manuf Eng 126:1–8
42. Rajurkar KP and Yu ZY (2000) Micro EDM can produce micro parts. Manuf Eng 125:1–4
43. Zheng HY and Huang H (2007) Ultrasonic vibration assisted femtosecond laser machining of
microholes. J Micromech Microeng 17:N58–N61
44. Tonshoff HK et al (1997) New machining concept for manufacturing microstructures with
excimer lasers. Mach Sci Technol 1:153–162
45. Gower MC (2000) Industrial application of laser micromachining. Opt Express 7:56–67
46. Weule H et al (2001) Micro-cutting of steel to met new requirements in miniaturization. Ann
CIRP 50/1:61–64
47. Egashira K and Masuzawa T (1999) Microultrasonic machining by the application of work-
piece vibration. Ann CIRP 48/1:131–134
48. Geiger M et al (2001) Microforming. Ann CIRP 50/2:445–462
49. Joo BY et al (2001) Development of micro punching system. Ann CIRP 50/1:191–194
50. Kurobe T et al (2001) Development of high speed slurry flow finishing of the inner wall of
stainless steel capillary – polishing and gas flow characteristics of various size of capillaries.
Precis Eng 25:100–106
51. Kurobe T et al (2002) Application of high speed slurry flow finishing method for finishing of
inner wall of fine hole die - effects of the hardness of die materials on the polishing
characteristics. Precis Eng 26:155–161
52. Sakuyama T et al (1999) Advances in Abrasive Jet Machining. Development of a new
abrasive jet machining device. In Wang J et al (ed) World Scientific, Singapore. 291–298
53. Yin L et al (2004) Abrasive flow polishing of micro bores. Mater Manuf Process 19:187–207
54. Giles RV and Pollard D (1997) Schaum’s outline of theory and problems of fluid mechanics
and hydraulics. McGraw-Hill, New York
55. Rhoades LJ (1987) Abrasive flow machining. Soc Manuf Eng MR87-163:1–15
56. Rhoades LJ (1991) Abrasive flow machining in a new era of manufacturing. Soc Manuf Eng
MR91-523:1–16
384 H. Huang et al.